Journal of Occupational Health Psychology © 2015 American Psychological Association

2016, Vol. 21, No. 1, 24 –36 1076-8998/16/$12.00 http://dx.doi.org/10.1037/a0039138

Distinct Longitudinal Patterns of Absenteeism and Their Antecedents in

Full-Time Australian Employees
Christopher A. Magee, Peter Caputi, and Jeong Kyu Lee
University of Wollongong

This paper investigated distinct longitudinal trajectories of absenteeism over time, and underlying
demographic, work, and health antecedents. Data from the Household, Income, and Labor Dynamics in
Australia Survey were used; this is a panel study of a representative sample of Australian households.
This paper focused on 2,481 full-time employees across a 5-year period. Information on annual sick leave
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

and relevant sociodemographic, work, and health-related factors was collected through interviews and
This document is copyrighted by the American Psychological Association or one of its allied publishers.

self-completed surveys. Growth mixture modeling indicated 4 distinct longitudinal patterns of absentee-
ism over time. The moderate absenteeism trajectory (34.8%) of the sample had 4 –5 days of sick leave
per year and was used as the reference group. The low absenteeism trajectory (33.5%) had 1 – 2 days of
absenteeism per year, while the no absenteeism trajectory (23.6%) had very low rates of absenteeism (⬍1
day per year). Finally, a smaller trajectory accounting for 8.1% of the sample had high levels of
absenteeism (⬎11 days per year). Compared with the moderate absenteeism trajectory, the high
absenteeism trajectory was characterized by poor health; the no absenteeism and low absenteeism
trajectories had better health but may also reflect processes relating to presenteeism. These results
provide important insights into the nature of absenteeism in Australian employees, and suggest that
different patterns of absenteeism over time could reflect a range of demographic, work, and health related
factors.

Keywords: absenteeism, growth mixture modeling, job security, work hours, health

Absenteeism, which refers to “the failure to report to work as
scheduled” (Johns, 2008, p. 160), is a major contributor to reduced
workplace productivity and has considerable financial costs
through health insurance claims, overtime wages, and legal claims
(Darr & Johns, 2008). Absenteeism can have longer term effects
on productivity by contributing to conditions such as depression
(Melchior et al., 2009) and leading to eventual withdrawal from
the workplace (Westman & Etzion, 2001). In Australia, average
rates of absenteeism range from 4 days per year (private sector
employees) to 8 days per year (public sector employees) and vary
by age, length of service, income, gender, organization size, and
industry type (Audit Office of New South Wales, 2010). It is
estimated that absenteeism costs the Australian economy at least
$7 billion a year in lost productivity and health care costs (Medi-
This article was published Online First May 4, 2015.
Christopher A. Magee, Centre for Health Initiatives, University of Wol-
longong; Peter Caputi, Centre for Health Initiatives and School of Psy-
chology, University of Wollongong; Jeong Kyu Lee, Centre for Health
Initiatives, University of Wollongong.
This paper uses unit record data from the Household, Income, and Labor
Dynamics in Australia (HILDA) Survey. The HILDA Project was initiated
and is funded by the Australian Government Department of Social Services
(DSS) and is managed by the Melbourne Institute of Applied Economic
and Social Research (Melbourne Institute). The findings and views re-
ported in this paper, however, are those of the authors and should not be
attributed to either DSS or the Melbourne Institute.
Correspondence concerning this article should be addressed to Christo-
pher A. Magee, Centre for Health Initiatives, University of Wollongong,
Wollongong NSW 2522, Australia. E-mail: cmagee@uow.edu.au
bank Private, 2005). Studies from North America and Europe also
indicate that the economic costs of absenteeism are considerable
(Confederation of British Industry, 2004; Cooper & Dewe, 2008;
Goetzel et al., 2004).
Although a vast body of literature has investigated antecedents
and consequences of absenteeism in numerous contexts (Johns,
2011), many aspects of absenteeism remain unclear. Few studies
have investigated whether there are distinct longitudinal patterns
of absenteeism in employees; this is an important consideration
because patterns of absenteeism over time could vary considerably
between individuals, reflecting the role of different antecedents.
The current paper aimed to investigate the presence of distinct
longitudinal trajectories of absenteeism, and their underlying an-
tecedents to better understand the nature of absenteeism. The
remainder of this introduction outlines some key theories of ab-
senteeism, and briefly reviews existing longitudinal research ex-
amining antecedents of absenteeism. The need to examine distinct
trajectories of absenteeism over time is then presented, followed
by the aims and main research questions of the present study.
Absenteeism Behavior
Absenteeism is a complex phenomenon, with many interrelated
factors having the potential to influence an employee’s decision
and ability to attend work. Historically, a distinction has been
drawn between involuntary absenteeism and voluntary absentee-
ism (Brooke, 1986; Johns, 2011; Nicholson, 1977; Steers & Rho-
des, 1978). Involuntary absenteeism (also known as sickness ab-
senteeism) refers to instances where employees do not attend work
due to poor physical or mental health (Brooke, 1986; Steers &
Rhodes, 1978). The term involuntary does not imply an employee

24
 ABSENTEEISM IN FULL-TIME AUSTRALIAN EMPLOYEES
lacks control over their decision to attend work; rather, it refers to
the influence of external factors in constraining an employee’s
ability to attend work (Beemsterboer, Stewart, Groothoff, & Ni-
jhuis, 2009). In general, involuntary absenteeism behavior is char-
acterized by duration rather than frequency of absenteeism spells
(Brooke, 1986). Voluntary absenteeism differs in that it reflects a
choice to withdraw or escape from a negative work environment
(Steers & Rhodes, 1978). In other words, employees may choose
to be absent from work as a way of escaping, avoiding, or com-
pensating for adverse, demoralizing, or stressful work environ-
ments (Hardy, Woods, & Wall, 2003). Voluntary absenteeism is
reflected by duration rather than frequency of missed days of work
(Brooke, 1986).
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Many conceptual models and frameworks have been proposed
This document is copyrighted by the American Psychological Association or one of its allied publishers.
to explain voluntary and involuntary absenteeism. It is beyond the
scope of this study to review all of these in detail; however some
key theoretical perspectives are briefly presented. Nicholson’s
(1977) attendance motivation model conceptualized absenteeism
behavior on a continuum from primarily unavoidable (involuntary)
to primarily avoidable (voluntary) absenteeism. Nicholson (1977)
argued that contextual factors (e.g., personal characteristics, job
characteristics, demographics) influence an employee’s level of
attachment to work, which in turn affects how motivated they are
to attend work. According to this model, attending work is the
norm behavior until proximal events interfere with an employee’s
ability to attend work or force them to make a decision about
whether to attend work. Nicholson (1977) suggested that employ-
ees with higher levels of attendance motivation are less affected by
these proximal events.
Steers and Rhodes (1978) proposed a widely cited conceptual
model of attendance at work that provides an insight into voluntary
and involuntary absenteeism. They proposed that demographic
characteristics have an indirect influence on attendance behavior
via two main variables: (a) an employee’s motivation to attend
work and (b) an employee’s ability to attend work. Motivation to
attend work reflects an employee’s affective response to their job
(e.g., job satisfaction) and a range of internal and external pres-
sures (e.g., economic, social, or personal pressures). According to
this model, when an employee enjoys his or her work, they have a
strong motivation to attend work, and thus take fewer absenteeism
days (Steers & Rhodes, 1978). The model also proposes that
factors such as illness and accidents, family responsibilities, and
transportation problems may inhibit an individual’s ability to at-
tend work regardless of motivation; these factors contribute to
involuntary absenteeism (Steers & Rhodes, 1978). This model was
extended by Brooke (1986) who outlined several exogenous (e.g.,
routinization, work involvement, role ambiguity, organizational
permissiveness) variables and five endogenous variables (job sat-
isfaction, job involvement, organizational commitment, health,
and alcohol consumption), and hypothesized pathways by which
these variables were related with each other and with absenteeism.
Nicholson and Johns (1985) proposed that in addition to many
person-level factors captured in the models outlined above, two
social factors - absence cultures and the psychological contract—
also have the potential to influence absenteeism. Absence cultures
in a workplace can directly influence absenteeism for a given
group of employees through shared norms. Employees may also
observe the absence behaviors of other employees and the reac-
tions to these behaviors (Nicholson & Johns, 1985). Furthermore,
25
absence cultures may facilitate or constrain the effects of person
level variables such as job satisfaction on absenteeism. Nicholson
and Johns (1985) specifically postulated that two interrelated fac-
tors influence absence culture: (a) beliefs about absence and (b)
assumptions about employment (the psychological contract).
Nicholson and Johns (1985) argued that absence culture and psy-
chological contract are able to account for differences in absen-
teeism behavior within and between organizations. For example,
factors such as occupational status (e.g., white collar vs. blue collar
workers) may lead to differing beliefs regarding the legitimacy of
absences, which translates to different absenteeism behaviors.
These and several more recent conceptual frameworks and mod-
els (Halbesleben, Whitman, & Crawford, 2014; Johns, 2010;
Schaufeli, Bakker, & Van Rhenen, 2009) provide insight into a
range of factors that influence absenteeism behavior along the
involuntary-voluntary continuum. They suggest that a range of
demographic, work and job related, social, and health factors have
the potential to influence absenteeism behaviors.
Existing Longitudinal Studies
A large number of studies have examined absenteeism behavior
and underlying antecedents however, few studies have been theory
based and most have been inconsistent in terms of the type and
scope of antecedents examined. This section does not intend to
provide an exhaustive review of the longitudinal evidence base,
which is vast and encompasses a wide range of antecedents.
Instead, the aim is to provide an indication of the main antecedents
of absenteeism, consistent with established theoretical frameworks
such as Steers and Rhodes (1978). For clarity, the findings are
organized according to demographic characteristics, job satisfac-
tion, psychosocial job characteristics, work characteristics, and
health.
Demographic Characteristics
Factors such as age, sex, and socioeconomic status (e.g., edu-
cation) have been found to predict levels of absenteeism. For
example, age is inversely associated with absenteeism (Labriola,
Lund, & Burr, 2006; Martocchio, 1989; Ng & Feldman, 2008),
while in terms of sex females tend to have higher rates of absen-
teeism compared with males (Labriola et al., 2006; Mastekaasa &
Olsen, 1998). Consistent with Steers and Rhodes (1978), these
characteristics may have direct or indirect effects on absenteeism
(e.g., indirect effects via health and motivation) or moderate the
effects of the variables noted below on absenteeism.
Job Satisfaction
Job satisfaction is regarded as a key motivational factor under-
lying absenteeism behavior. In general, employees who have
higher levels of job satisfaction have lower rates of absenteeism
(Schaufeli et al., 2009; Ybema, Smulders, & Bongers, 2010). This
is consistent with the proposition of many existing frameworks
that employees who are more satisfied with aspects of their jobs
will be more motivated to attend work and thus have lower rates of
absenteeism (e.g., Brooke, 1986; Steers & Rhodes, 1978).
 26 MAGEE, CAPUTI, AND LEE
Psychosocial Job Characteristics
Many longitudinal studies utilizing theories such as the job-
demands control model (Karasek, 1979), job demands resources
model (Schaufeli et al., 2009), and the effort-reward imbalance
model (Siegrist, 1996) have demonstrated that aspects of the
psychosocial work environment are associated with absenteeism
behavior. For example, low job control (Nielsen, Rugulies, Chris-
tensen, Smith-Hansen, & Kristensen, 2006; Smulders & Nijhuis,
1999; Väänänen et al., 2003; Virtanen et al., 2007), high job strain
(Suominen et al., 2007), low social support (Melchior, Niedham-
mer, Berkman, & Goldberg, 2003), poor role clarity (Rugulies et
al., 2007), workplace bullying (McTernan, Dollard, & LaMon-
tagne, 2013), reduced organizational commitment (Bakker, De-
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
merouti, de Boer, & Schaufeli, 2003), poor organizational climate,
This document is copyrighted by the American Psychological Association or one of its allied publishers.
and poor leadership (Rugulies et al., 2007) have all been linked
with higher absenteeism. Findings for factors such as job demands
have been mixed (Smulders & Nijhuis, 1999; Virtanen et al.,
2007), with some studies finding that higher job demands predict
absenteeism in females but not males (Nielsen et al., 2006; Vir-
tanen et al., 2007). Some studies have found that changes in
psychosocial characteristics over time (e.g., reductions in job con-
trol and increases in job demands) predict greater absenteeism
(Head et al., 2006; Vahtera, Kivimäki, Pentti, & Theorell, 2000).
According to the theoretical perspectives outlined in the previous
section, these psychosocial factors may lead to more absenteeism
via lower job satisfaction and motivation to attend work, or by
hindering the ability to attend work through elevated stress and
poor health (Brooke, 1986; Nicholson, 1977; Steers & Rhodes,
1978).
Studies have increasingly examined the role of job security in
absenteeism, and have revealed mixed results (Blekesaune, 2012).
Some studies report that lower levels of job security lead to more
absenteeism (Kivimäki et al., 1997); this can be interpreted in the
context of stressor perspectives, as lower job security could lead to
greater psychological distress and hence more sick leave (Bleke-
saune, 2012). In contrast, other studies have found that lower
levels of job security lead to less absenteeism (Blekesaune, 2012);
this finding can be explained by a disciplinary effect, as lower job
security may be associated with fear of losing one’s job which
leads to less absenteeism (Blekesaune, 2012). Blekesaune (2012)
suggested that the disciplinary effect is most likely to account for
short-term absenteeism, while the stressor effect accounts for
longer-term absenteeism.
Work Characteristics
Characteristics of an individual’s work such as their job type,
work schedule, and work hours have been linked with levels of
absenteeism. For example, white-collar workers have been found
to have lower absenteeism compared with blue-collar workers
(Pousette & Hanse, 2002), which is consistent with Nicholson and
Johns’ (1985) model. Studies have shown that working night shifts
is associated with increased absenteeism; for many employees
shift work is demanding and can be a source of stress, which could
lead to poorer health and thus higher absenteeism (Fekedulegn et
al., 2013). Findings for work hours are not clear, although some
studies indicate that longer work hours are associated with lower
absenteeism (Magee, Stefanic, Caputi, & Iverson, 2011). Other
factors such as no access to sick leave benefits or self-employed
status are also associated with lower rates of absenteeism (Bena-
vides, Benach, Diez-Roux, & Roman, 2000).
Health and Illness
Health-related factors have consistently been shown to predict
absenteeism behavior (Labriola et al., 2006; Schalk, 2011; ten
Brummelhuis, Ter Hoeven, De Jong, & Peper, 2013). Short-term
absences from work are often attributed to acute illnesses such as
the common cold and influenza (Schaufeli et al., 2009), whereas
long-term absences are attributed to chronic mental and physical
health conditions including pain (e.g., neck pain), long-term dis-
ability, hypertension, depression, and migraines (Kääriä, Laak-
sonen, Leino-Arjas, Saastamoinen, & Lahelma, 2012; Schaufeli et
al., 2009). Poor psychological health (e.g., depression, anxiety,
general psychological distress) and sleep disturbances (e.g., insom-
nia) also predict absenteeism (Knudsen, Harvey, Mykletun, &
Øverland, 2013; Roelen et al., 2014; Wada et al., 2013). Consistent
with Steers and Rhodes (1978), poor health may contribute to
involuntary absenteeism by constraining one’s ability to attend
work.
Trajectories of Absenteeism Behavior
A large body of longitudinal research therefore indicates that a
range of factors have the potential to influence absenteeism be-
havior. However, these studies provide only a partial insight into
the nature of absenteeism and its antecedents. This is because most
studies utilize analytic techniques that do not capture potential
interindividual and intraindividual differences in absenteeism over
time. These are important considerations as patterns of absentee-
ism behavior may vary considerably across time and between
employees. These variations could be reflected in distinct longi-
tudinal trajectories of absenteeism, which may have distinct ante-
cedents. For example, some absenteeism trajectories could reflect
the influence of motivational factors, some may reflect more
constraint factors (e.g., health), while others reflect a combination
of factors. Investigation of these intraindividual and interindividual
variations in absenteeism, along with their antecedents, is needed
to provide a more definitive insight into the nature of absenteeism.
To date, only a small number of studies have explored distinct
longitudinal trajectories of absenteeism (Dello Russo, Miraglia,
Borgogni, & Johns, 2013; Haukka et al., 2014; Haukka et al.,
2013). In a representative sample of 3,420 Finnish employees
followed for 7 years, Haukka and colleagues (2013) identified four
distinct trajectories of absenteeism spells (defined as sickness
absences periods spanning 10 or more consecutive working days).
The largest trajectory (“low” sick leave) included 59% of employ-
ees who had no occurrences of long sickness absence spells. A
small proportion of participants (9%) had a high occurrence of
sickness absences each year. The remaining two trajectories were
intermediary groups with ascending and descending patterns of
absenteeism occurrences. Factors such as increased age, poor
health (e.g., musculoskeletal, mental illness, obesity), physical
workload, and low job control predicted the higher sick leave
trajectories, suggesting that health and motivational factors could
underlie these trajectories. In a subsequent study of female kitchen
workers, Haukka et al. (2014) found three distinct absenteeism
trajectories over a 2-year period: no absenteeism; intermediate
 ABSENTEEISM IN FULL-TIME AUSTRALIAN EMPLOYEES
absenteeism, and high absenteeism. Bodily pain, smoking, and
obesity predicted the intermediate trajectory, while depression,
musculoskeletal disease, and bodily pain predicted the high absen-
teeism trajectory. These findings suggest that different health
factors have unique influences on patterns of absenteeism.
Dello Russo at al. (2013) adopted a different approach and
examined trajectories of absenteeism behavior in three groups of
employees on the basis of how long they had been at the organi-
zation (i.e., their duration of tenure). The three groups examined
were short-tenured (tenure ⬍3 years), medium-tenured (tenured
between 3 and 19 years), and high-tenured employees (tenure ⬎19
years). At baseline, short tenured employees showed low rates of
annual absenteeism (average 2 days a year) compared with the
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
other two groups. However, this group showed an increase in
This document is copyrighted by the American Psychological Association or one of its allied publishers.
absenteeism behavior over a 4-year period, such that levels of
absenteeism increased to 4.5 days per year, which was similar with
medium tenured employees. The results were interpreted within
the context of organizational norms. That is, employees new to an
organization initially showed lower rates of absenteeism, but a
gradual increase in absenteeism to be consistent with those with
longer job tenure. These results suggest that employees gradually
conform to the dominant norm of the organization in relation to
absenteeism behavior (Dello Russo, et al., 2013).
The Present Study
It has been well demonstrated that absenteeism is an important
indicator of workplace productivity. Furthermore, several concep-
tual frameworks and models have identified a number of anteced-
ents of absenteeism behavior, which have subsequently been sup-
ported by longitudinal research. However, a key limitation of
existing studies is that intraindividual and interindividual varia-
tions in absenteeism over time have rarely been captured in pre-
vious studies. This is an important consideration because there
may be distinct longitudinal trajectories of absenteeism, which
have different underlying antecedents.
The aim of the present study was to investigate the presence and
nature of absenteeism in a sample of full-time Australian employ-
ees over a 5-year period. The present study involved using a
growth mixture modeling approach to identify and examine dis-
tinct longitudinal trajectories of absenteeism and their antecedents.
This involved examining the following primary research question:
Research Question 1: Are there distinct longitudinal trajecto-
ries of absenteeism in full-time employees?
We also aimed to examine antecedents that predicted and dis-
tinguished between these trajectories. Based on the widely cited
Steers and Rhodes (1978) model and existing empirical research,
we focused on antecedents reflecting demographic characteristics
(e.g., age, sex, education), job satisfaction, psychosocial work
characteristics (e.g., job demands, job control, job security), work
characteristics (e.g., work hours, work schedules, job type, sick
leave entitlements), and health. This involved addressing three
additional research questions.
Research Question 2: Do demographic characteristics (age,
sex, education) predict absenteeism trajectories?
27
Research Question 3: Are characteristics of an individual’s
job (e.g., work characteristics and psychosocial job charac-
teristics) associated with absenteeism trajectories?
Research Question 4: Does health status distinguish between
absenteeism trajectories?
Method
Participants
The Household, Income, and Labor Dynamics in Australia
(HILDA) Survey is a representative longitudinal panel study of
Australian households (Wooden, Freidin, & Watson, 2002).
HILDA, which commenced in 2001, collects data from members
of Australian households every 12 months through a series of
interviews and self-completion surveys. In the first Wave of the
study, members from 11,693 Australian households were invited
to participate in the study. Of those households contacted, 7682
provided data from at least one household member leading to an
initial sample size of 19,910. Of these individuals, 4,787 were aged
younger than 15 years at the time of interview and were excluded,
leaving a sample size of 13,158.
This paper utilized data from five recent waves of data (Waves
7 to 11). In this paper, we included participants aged 18 years and
older who were employed full-time at each wave (n ⫽ 2,934);
participants with missing absenteeism data were excluded (n ⫽
453). This resulted in a final sample of 2,481 full-time employees.
Table 1 shows the characteristics of the final sample compared
with those with missing absenteeism data. This table indicates
some differences between the samples, particularly in terms of
health and sick leave benefits. The HILDA Survey received ethics
approval from the University of Melbourne Human Research
Ethics Committee. Ethical approval to use these data in the present
paper was obtained from our university’s Human Research Ethics
Committee.
Measures
Absenteeism. Absenteeism was measured in relation to two
interview questions that asked participants whether they had taken
any sick leave in the previous 12 months, and if so, the number of
sick leave days they had taken. The data were combined to create
a count indicator of the amount of annual sick leave (in days) taken
over the 5-year period.
Demographic factors. Information on age, gender, and high-
est level of education (coded as high school, completed high
school, diploma/certificate, or university degree) was collected and
included in the analysis.
Psychosocial work characteristics. The HILDA Survey con-
sisted of five items examining different aspects of job satisfaction,
including satisfaction with job security, hours of work, and work
life balance. These items were assessed on an 11-point scale, with
higher scores on all items indicative of higher levels of satisfac-
tion.
Participants completed 12 items that examined aspects of their
perceived work environment (e.g., “I have to work fast in my job”;
“I have a lot of say in what I do at work”). Each item was assessed
on a 7-point Likert scale (strongly disagree to strongly agree).
 28 MAGEE, CAPUTI, AND LEE

Table 1
Characteristics of the Sample Compared With Those Excluded Due to Missing Data

Excluded due to
Final sample missing data p Value for
(n ⫽ 2,481) (n ⫽ 453) differencea

Sex, n (%) .791

Female 735 (84.7) 137 (15.7)
Male 1746 (84.3) 316 (15.3)
Age, mean (SD) 39.75 (11.06) 39.47 (12.24) .621
Bodily Pain subscale, mean (SD) 79.80 (19.39) 74.48 (20.83) .005
General Health subscale, mean (SD) 73.56 (17.54) 69.02 (18.92) .005
Satisfaction—job security, mean (SD) 8.28 (11.80) 8.05 (1.94) .014
Satisfaction—work hours, mean (SD) 7.13 (1.92) 6.98 (1.99) .135
Job control, mean (SD) 26.34 (8.61) 27.63 (9.68) .166
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Work Schedule, n (%) .272

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Nonstandard 408 (82.9) 369 (15.1)

Standard 2,073 (84.9) 84 (17.1)
Job type, n (%) .036
Professional 428 (81.5) 97 (18.5)
Manager 667 (86.7) 102 (13.3)
Trade/Technician 432 (82.3) 93 (17.7)
Laborer 629 (86.3) 100 (13.7)
Clerical/Administration/Sales 325 (84.2) 61 (15.8)
Sick leave benefits, n (%) ⬍.001
No benefits 139 (85.3) 24 (14.7)
Self-employed/Business owner 334 (75.2) 110 (24.8)
Access to benefits 2008 (86.3) 319 (13.7)
Work hours, n (%) .020
35–39 hr 547 (84.3) 102 (15.7)
40–49 hr 1,173 (86.4) 185 (13.6)
50 hr or more 761 (82.1) 166 (17.9)
a
p Values derived from chi-square tests (categorical variables) and ANOVAs (continuous variables).

Exploratory and confirmatory factor analysis conducted in this
paper indicated that there were two distinct factors. Factor 1
assessed issues surrounding flexibility and autonomy at work and
was consequently labeled job control (Cronbach’s alpha ⫽ .89).
Factor 2 assessed intensity of work and time pressures and was
labeled job demands (Cronbach’s alpha ⫽ .77). Higher scores are
indicative of high levels of job control and job demands respec-
tively.
Work characteristics. The interviews and self-report ques-
tionnaires also collected information on the number of hours
worked each week (coded as 35–39 hr, 40 – 49 hr, and ⱖ50 hr a
week) and work schedule (coded as standard work hours [i.e.,
regular daytime shift] and nonstandard work hours [evening/night
shifts, rotating shifts, irregular schedules]). Information was also
collected on the type of job according to the Australian and New
Zealand Standard Classification of Occupations (Australian Bu-
reau of Statistics, 2006). This classification system distinguishes
between eight major groupings of job types: (a) managers, (b)
professionals, (c) technicians and trades workers, (d) community
and personal service workers, (e) clerical and administrative work-
ers, (f) sales workers, (g) machinery operators and drivers, and (h)
laborers. Due to some small cell sizes, we merged Categories d, e,
and f into a single category and g and h into a single category. This
resulted in five job type categories.
Health. Self-reported mental and physical health were as-
sessed via the Short-Form Health Survey (SF-36), a 36-item scale
that assesses health across eight subscales (Ware, Kosinski, &
Gandek, 1993, 2000). The Physical Functioning subscale (10
items) examines how well individuals are able to perform normal
activities of daily living such as climbing stairs and kneeling. The
Role Physical subscale (four items) assesses the extent to which
physical health affects problems with work and other daily activ-
ities. The Bodily Pain subscale includes two items regarding the
extent to which individuals experience pain and the effect on
subsequent daily activities. Social Functioning (two items) asks
participants to indicate the extent to which their physical and
emotional problems affect their normal daily activities. Vitality
(four items) assesses individual’s levels of vigor and energy. Role
Emotional (three items) examines the effect of emotional problems
on work or daily activities. Finally, the mental health subscale
includes five items that assess levels of depression and anxiety. On
all subscales, higher scores are indicative of better health.
Statistical Analysis
Longitudinal trajectories of absenteeism were investigated using
growth mixture modeling (GMM), which is a statistical approach
that examines change in a given variable over time (Jung &
Wickrama, 2008). Conventional growth modeling assumes that a
single growth curve (or trajectory) is adequate for capturing
change in an entire population, and that covariates affect growth
factors in the same way for all individuals (Jung & Wickrama,
2008). However, there are often distinct subpopulations within a
sample that have different patterns of change over time (Jung &
Wickrama, 2008); these distinct patterns or trajectories are not
accounted for using conventional analytic approaches. This is
 ABSENTEEISM IN FULL-TIME AUSTRALIAN EMPLOYEES 29

particularly relevant in the context of absenteeism given some Results

recent findings suggesting identified distinct longitudinal patterns
of absenteeism (Dello Russo, et al., 2013; Haukka et al., 2014; Table 2 shows the model fit results for models tested in Step 1.
These results indicate that specifying additional trajectories (e.g.,
Haukka et al., 2013). In addition, it is feasible that potential
two vs. one trajectory) led to improvements in model fit as indi-
antecedents of absenteeism have a differential effect on these
cated by significant BLRT results and lower relative values for the
trajectories, which can also be examined using GMM. Therefore,
AIC, BIC, and sample-size–adjusted BIC. Despite these improve-
GMM is an ideal approach for the present study because it can
ments, the size of each additional trajectory identified became
identify distinct trajectories of absenteeism as well as covariates
smaller. In the four-class model, the smallest trajectory accounted
that distinguish between the trajectories. Because absenteeism data
for approximately 8% of the sample, while in models with five or
were counts, with a highly skewed distribution and a high propor-
more trajectories, the smallest trajectory accounted for ⬍3% of the
tion of zero values, zero-inflated Poisson GMMs were tested in
sample. These trajectories may be too small to examine meaning-
this paper (Muthén & Muthén, 1998 –2010).
fully. As a consequence, the four-trajectory model was identified
Consistent with recent recommendations, the analyses involved
as the optimal solution, which is consistent with the number
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

three main steps (Jung & Wickrama, 2008; Muthén, 2006; Muthén, identified by Haukka et al. (2013, 2014).
This document is copyrighted by the American Psychological Association or one of its allied publishers.

2004). The first step, performed using Mplus Version 6.11 In the second stage of the analyses, the multinomial logistic
(Muthén & Muthén, 1998 –2010), aimed to identify the number of regression models conducted with SPSS Version 19 indicated that
distinct trajectories of absenteeism. This step involved testing a age, sex, education, job type, work hours, work schedules, sick
model with a single trajectory, following by a model with two leave entitlements, job control, satisfaction with job security, sat-
trajectories, then three trajectories and so on until the optimal isfaction with hours worked, general health, and bodily pain were
number of trajectories was identified. For this first step, covariates significantly associated with class membership as specified by the
were not included in the models. Several sources of information four-class model.
were used to decide on the optimal number of trajectories. Three The full GMM was then tested in Mplus, with a four-trajectory
information criteria—Akaike information criteria (AIC), Bayesian model specified and the significant covariates from Step 2 in-
information criteria (BIC), and sample-size–adjusted BIC—were cluded as time-invariant covariates in the model. The four absen-
used to compare sequential models, with lower values for a model teeism trajectories resulting from the full model are shown in
with k trajectories (e.g., three trajectories) suggesting an improved Figure 1. The first trajectory (n ⫽ 879; 35.3%) included employees
model fit compared with a model with k – 1 trajectories (e.g., two who took 4 to 5 days of sick leave annually. The amount of
trajectories). Relying on these criteria alone can overestimate the absenteeism increased over time at a decreasing rate as reflected
number of trajectories, so bootstrap likelihood ratio tests (BLRT) by the significant linear (B ⫽ .18, p ⬍ .001) and quadratic
were also used (Nylund, Asparouhov, & Muthén, 2007). The functions (B ⫽ ⫺.03, p ⫽ .038). This trajectory was labeled
BLRT compares model fit between subsequent models (e.g., com- moderate absenteeism.
pares a three class model with a two class model); a significant The second trajectory included 8.4% (n ⫽ 210) of the sample,
BLRT suggests that the model with one more trajectory provides and had high levels of absenteeism compared with the other
a significant improvement in model fit (Nylund et al., 2007). The trajectories. The linear (B ⫽ ⫺.01, p ⫽ .891) and quadratic
optimal number of trajectories is informed by the specification of functions (B ⫽ .02, p ⫽ .573) were not significant for this
additional trajectories not leading to a significant BLRT. In decid- trajectory, indicating no significant changes in absenteeism over
ing on the optimal number of latent classes, we additionally time. This trajectory was labeled high absenteeism.
considered classification accuracy (informed by entropy ⬎ .80), The third trajectory comprised 23.3% (n ⫽ 579) of the present
trajectory size (to ensure trajectories were large enough to be sample, and included employees who essentially took no sick leave
examined meaningfully), and the distinctiveness of trajectories (to
ensure identified trajectories were distinct from one another).
The aim of Step 2 was to identify covariates that were signifi- Table 2
cantly associated with the trajectories identified in Step 1. This Model Fit Results for Growth Mixture Models Specifying 1 to
involved conducting multinomial logistic regression modeling using 8 Classes
SPSS Version 19 to investigate covariates significantly associated
Sample-size– BLRT
with trajectory membership. Covariates that were significantly asso- Classes AIC BIC adjusted BIC p value Entropy
ciated with the trajectories were retained for inclusion in Step 3.
The third step of the analysis involved testing a full GMM, 1 91,038.90 91,056.29 91,046.76 — 1.00
2 69,405.93 69,446.49 69,424.25 ⬍.001 .94
whereby the significant covariates from Step 2 were added as
3 63,474.33 63,538.06 63,503.11 ⬍.001 .94
time-invariant covariates to the GMM (specifying the number of 4a 61,534.55 61,621.41 61,573.75 ⬍.001 .91
distinct trajectories from Step 1). This step allowed the antecedents 5 59,947.14 60,057.22 59,996.86 ⬍.001 .91
to influence the growth factors of the trajectories, and provided 6 58,740.55 58,873.82 58,800.74 ⬍.001 .92
7 57,845.84 58,002.28 57,916.49 ⬍.001 .93
insight into how the trajectories varied by these antecedents
8 57,022.71 57,202.33 57,103.83 ⬍.001 .92
(Muthén, 2004). This final model thus indicated the nature of the
different trajectories and significant covariates that distinguished Note. AIC ⫽ Akaike’s information criteria; BIC ⫽ Bayesian information
criteria; BLRT ⫽ bootstrap likelihood ratio test.
between these trajectories. The differences between the trajectories a
The four-class model was considered optimal, given that models with five
in regards to the antecedents are reported as odds ratios based on or more classes identified very small trajectories, despite improvements in
logistic regressions conducted within the GMM. model fit.
 30 MAGEE, CAPUTI, AND LEE
No sick leave
Low sick leave
Moderate sick leave
High sick leave
15
14
13
Number of sick leave days
12
11
10
9
8 p ⬍ .05), and poorer health as reflected by lower scores on the
7 Bodily Pain (OR ⫽ .99, p ⬍ .05), and General Health (OR ⫽ .99,
6 p ⬍ .05) subscales.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
5
4
This document is copyrighted by the American Psychological Association or one of its allied publishers.
3 Discussion
2
1 This study utilized a GMM approach to identify distinct trajec-
0
2007 2008 2009 2010 2011
Years
Figure 1. The four distinct trajectories of annual absenteeism in the
sample of Australian full-time employees. Means and standard deviations
are shown for each trajectory.
annually across the 5-year period. There was a significant trend for
a decline in absenteeism over time as reflected by the linear
function (B ⫽ ⫺.78, p ⫽ .003); the quadratic function indicated
that this decline slowed with time (B ⫽ .19, p ⫽ .013). This
trajectory was labeled no absenteeism.
The final trajectory included 33.0% (n ⫽ 822) of the present
sample, and included employees who took 1 to 2 days of sick leave
over the 5-year period; levels of absenteeism remained stable over
time as reflected by nonsignificant linear (B ⫽ ⫺.05, p ⫽ .552)
and quadratic growth functions (B ⫽ .01, p ⫽ .709). This trajec-
tory was labeled low absenteeism.
The characteristics of these trajectories are shown in Table 3
with the multivariate differences between the trajectories shown in
Table 4. In order to interpret the multivariate differences in these
characteristics between trajectories, it was necessary to select one
trajectory as the referent. We decided on the moderate absenteeism
trajectory as the referent for two main reasons. First, it was the
largest trajectory, which can be used as a justification for selecting
a suitable reference category. Second, the pattern of sick leave
corresponds most closely with average levels of sick leave in
Australia. Therefore, it may provide a useful indication of “typi-
cal” absenteeism patterns, and allow for meaningful comparisons
with the other trajectories.
The GMM indicated that several covariates distinguished be-
tween the trajectories (see Table 4). Compared with the moderate
absenteeism trajectory, the no absenteeism trajectory had a lower
proportion of females (odds ratio [OR] ⫽ .47, p ⬍ .001), were
younger (OR ⫽ .96, p ⬍ .001), had better general health (OR ⫽
1.02, p ⬍ .001), and higher levels of job control (OR ⫽ 1.03, p ⬍
.05). In addition, individuals in the no absenteeism trajectory were
more likely to be managers (OR ⫽ 2.01, p ⬍ .001), self-employed/
business owner (OR ⫽ 39.65, p ⬍ .001), and work long hours
(OR ⫽ 2.01, p ⬍ .001) compared with moderate absenteeism
employees.
Compared with the moderate absenteeism trajectory, the low
absenteeism trajectory had a lower proportion of females (OR ⫽
.50, p ⬍ .001), were older (OR ⫽ 1.01, p ⬍ .05), and had higher
general health (OR ⫽ 1.02, p ⬍ .001). These employees were also
more likely to be a manager (OR ⫽ 1.58, p ⬍ .05), self-employed/
business owner (OR ⫽ 3.72, p ⬍ .001), and work long hours
(OR ⫽ 1.73, p ⬍ .001) compared with moderate absenteeism
employees.
Compared with the moderate absenteeism trajectory, the high
absenteeism trajectory was characterized by older age (OR ⫽ 1.03,
tories of absenteeism and their antecedents in a sample of full-time
Australian employees. The results identified four distinct longitu-
dinal trajectories of absenteeism, which were labeled no absentee-
ism (23.3%), low absenteeism (33.0%), moderate absenteeism
(35.3%), and high absenteeism (8.4%). A small proportion of
individuals displayed annual fluctuations in absenteeism that de-
viated from these trajectories, perhaps reflecting factors such as
injury, short-term illness, or organization changes that produce
short-term changes in absenteeism (Hansson, Vingård, Arnetz, &
Anderzén, 2008). However, in general, the mean levels of absen-
teeism within each trajectory remained fairly stable over time. The
present findings add to a small number of existing studies that have
identified distinct trajectories of absenteeism behavior (Dello
Russo, et al., 2013; Haukka et al., 2013, 2014).
A key strength of the present study is that potential antecedents
of these trajectories were examined. The antecedents explored in
this study were consistent with well-established theories of absen-
teeism, most notably Steers and Rhodes’ (1978) model which
proposes that absenteeism behavior reflects demographic charac-
teristics, motivation to attend work, and ability to attend work. As
will be discussed in more detail, our results indicate that the
absenteeism trajectories differ on a range of these antecedents,
suggesting differences in motivation and ability to attend work. In
discussing these results, we utilize the moderate absenteeism tra-
jectory as the reference trajectory. This is because this trajectory
was the largest and had rates of absenteeism (4 – 6 days per year)
that corresponded most closely with average absenteeism rates in
Australia. Therefore, it provides a meaningful point of comparison
for the other trajectories. The findings are discussed separately for
each absenteeism trajectory.
High Absenteeism Trajectory
The high absenteeism trajectory was the smallest of the four
trajectories and was characterized by 11–13 days of absenteeism
each year. There were several antecedents that distinguished the
high absenteeism trajectory from the moderate absenteeism trajec-
tory. In particular, employees in the high absenteeism trajectory
had lower scores on the Bodily Pain and General Health subscales
of the SF-36, which indicates poorer health. The General Health
subscale assesses overall health, the likelihood of getting sick
relative to others, and whether the individual expects their health
 ABSENTEEISM IN FULL-TIME AUSTRALIAN EMPLOYEES 31

Table 3
Characteristics of the Four Sick Leave Trajectories Based on the Covariates Included in the Full Growth Mixture Model

No absenteeism Low absenteeism Moderate absenteeism High absenteeism

Sick leave, mean (SD)

Year 1 .16 (.52) 1.92 (2.36) 4.25 (3.46) 11.08 (11.22)
Year 2 .09 (.37) 1.95 (1.98) 5.01 (3.87) 11.09 (10.12)
Year 3 .07 (.33) 1.85 (1.90) 5.34 (3.80) 11.06 (10.92)
Year 4 .05 (.24) 1.69 (1.67) 5.63 (4.58) 12.30 (12.40)
Year 5 .13 (.50) 1.91 (2.10) 5.31 (4.04) 13.00 (11.62)
Age mean (SD) 42.94 (10.56) 39.29 (11.09) 37.64 (10.95) 41.58 (10.63)
Sex, n (%)
Male 487 (27.9) 626 (35.9) 526 (30.1) 107 (6.1)
Female 93 (12.7) 195 (26.5) 347 (47.2) 100 (13.6)
Bodily Pain subscale, mean (SD) 80.39 (18.28) 82.56 (18.17) 79.09 (19.51) 70.17 (22.98)
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

General Health subscale, mean (SD) 75.44 (16.92) 75.92 (16.58) 72.01 (17.42) 65.60 (20.21)
This document is copyrighted by the American Psychological Association or one of its allied publishers.

Job security, mean (SD) 8.13 (1.92) 8.31 (1.72) 8.30 (1.83) 8.49 (1.68)
Satisfaction with work hours, mean (SD) 6.80 (2.05) 7.15 (1.89) 7.37 (1.81) 6.92 (1.92)
Job control, mean (SD) 30.37 (8.04) 25.92 (8.21) 24.76 (8.47) 23.37 (8.52)
Work schedule, n (%)
Standard 455 (21.9) 707 (34.1) 732 (35.3) 179 (8.6)
Nonstandard 125 (30.6) 114 (27.9) 141 (34.6) 28 (6.9)
Job type, n (%)
Professional 165 (38.6) 151 (35.3) 92 (21.5) 20 (4.7)
Manager 115 (17.2) 220 (33.0) 259 (38.8) 73 (10.9)
Trade/Technician 122 (28.2) 143 (33.1) 144 (33.3) 23 (5.3)
Laborer 101 (16.1) 196 (31.2) 264 (42.0) 68 (10.8)
Clerical, etc. 77 (23.7) 111 (34.2) 114 (35.1) 23 (7.1)
Sick leave benefits, n (%)
No benefits 71 (51.1) 42 (30.2) 24 (17.3) 2 (1.4)
Self-employed 263 (78.7) 55 (16.5) 12 (3.6) 4 (1.2)
Access to benefits 246 (12.3) 724 (36.1) 837 (41.7) 201 (10.0)
Work hours, n (%)
35–39 hr 61 (11.2) 152 (27.8) 262 (47.9) 72 (13.2)
40–49 hr 215 (18.3) 401 (34.2) 458 (39.0) 99 (8.4)
50 hr or more 304 (39.9) 268 (35.2) 153 (20.1) 36 (4.7)

to get worse over time (Ware et al., 1993, 2000). Although devel-
oped as a measure of physical health, research has consistently
demonstrated that the General Health subscale also reflects aspects
of mental health (Ware et al., 1993, 2000). The Bodily Pain
subscale of the SF-36 assesses the levels of pain experienced in the
past 4 weeks, and the extent to which pain interfered with work
and housework. In combination, these two results indicate that
employees in the high absenteeism trajectory experienced more
pain and poorer physical and mental health compared with the
moderate absenteeism trajectory. Consistent with Steers and Rho-
des (1978), as well as many other theoretical perspectives (e.g.,
Brooke, 1986; Nicholson, 1977), it is feasible that poor health and
pain observed in this trajectory lead to higher rates of absenteeism
by constraining employee’s ability to attend work. These findings
suggest that high rates of absenteeism may primarily reflect health
constraint factors.
Low Absenteeism Trajectory
The low absenteeism trajectory accounted for one third of the
sample, with average rates of absenteeism in this trajectory ranging
from 2–3 days per year. Several characteristics distinguished em-
ployees in this trajectory from those in the moderate absenteeism
trajectory. First, employees in the low absenteeism trajectory were
older and more likely to be male. These findings are consistent
with existing studies demonstrating that older age and male gender
are associated with lower amounts of absenteeism (Darr & Johns,
2008; Feeney, North, Head, Canner, & Marmot, 1998; Martocchio,
1989; Mastekaasa & Olsen, 1998; Ng & Feldman, 2008; Steers &
Rhodes, 1978). The reasons for different rates of absenteeism in
males and females are not clear. Some existing studies have
suggested that a range of factors including differences in job type
and child care responsibilities could underlie these gender differ-
ences (Darr & Johns, 2008; Mastekaasa & Olsen, 1998), but this
requires further investigation using measures that are able to
examine these issues. In regards to age, existing research suggests
that older employees have lower rates of absenteeism because they
are better able to meet the requirements of their jobs (Martocchio,
1989) and/or have higher levels of job motivation (Ng & Feldman,
2008).
The results also indicated that employees in the low absenteeism
trajectory had better health compared with the moderate absentee-
ism trajectory as reflected by higher scores on the General Health
subscale of the SF-36. Consistent with Steers and Rhodes (1978),
better mental and physical health could mean that an individual is
more able to attend work, and have lower rates of absenteeism.
Individuals in the low absenteeism trajectory were also more
likely to be managers, work long hours, and be self-employed/a
business owner. Previous studies have indicated differences in
absenteeism rates by occupational status and job level (Darr &
Johns, 2008; Nicholson & Johns, 1985). For example, white collar
 32 MAGEE, CAPUTI, AND LEE

Table 4
Logistic Regression Results From the Full Growth Mixture Model

No absenteeism Low absenteeism High absenteeism

Sex
Female 0.47ⴱⴱ .50ⴱⴱ 1.19
Male Ref Ref Ref
Age 0.96ⴱⴱ 1.01ⴱ 1.03ⴱ
Bodily Pain subscale 1.01 1.01 0.99ⴱ
General Health subscale 1.02ⴱⴱ 1.02ⴱⴱ 0.99ⴱ
Job Satisfaction—job security 1.02 .99 1.11
Job Satisfaction—work hours 0.94 .96 0.91
Job control 1.03ⴱ 1.00 .98
Work schedule
Nonstandard 1.34 .82 0.68
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Standard Ref Ref Ref

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Job type
Professional .75 .94 1.10
Manager 2.01ⴱⴱ 1.58ⴱ .85
Trade/Technician 1.35 1.04 .80
Laborer .97 .95 .83
Clerical/Administration/Sales Ref Ref Ref
Sick leave benefits
No benefits 2.39 1.39 1.36
Self-employed/Business owner 39.65ⴱⴱ 3.72ⴱⴱ 2.56
Access to benefits Ref Ref Ref
Work hours
35–39 hr Ref Ref Ref
40–49 hr 1.13 .81 0.77
50 hr or more 2.01ⴱⴱ 1.73ⴱⴱ 0.80
Note. Ref ⫽ referent category. The low absenteeism trajectory is the referent category. Results are presented
as odds ratios.
ⴱ
p ⬍ .05. ⴱⴱ p ⬍ .001.

workers such as managers and professionals have been shown to
have lower rates of absenteeism compared with blue collar work-
ers (Darr & Johns, 2008; Nicholson & Johns, 1985). Although we
are unable to determine the reasons for these differences from the
present data, Nicholson and Johns’ (1985) work on absence culture
and psychological contract suggests that certain job types (including
managers) have lower rates of absenteeism because of guilt over
absences and a perception of absences as illegitimate (Nicholson &
Johns, 1985). Thus, some employees in managerial roles may avoid
taking time off work when needed because of these factors.
Employees in this trajectory were also more likely to be self-
employed or a business owner. Although all Australian employees,
with the exception of those employed casually, are eligible for paid
sick leave (Fair Work Ombudsman, 2010), individuals who are self-
employed may be reluctant to take time of work due to factors such
as illness because of potential for lost income (Benavides et al., 2000).
Individuals in this trajectory were also more likely to work longer
hours, which is consistent with some findings demonstrating that
longer work hours are linked with less absenteeism (Magee et al.,
2011).
The results discussed above suggest a range of different pro-
cesses underlying the low levels of absenteeism in this trajectory.
On the one hand, it is feasible that because of their better general
health relative to the moderate absenteeism trajectory these em-
ployees do not need to take as many days off work each year. That
is, for these employees better general health means they are less
constrained to attend work. Thus, this could be a healthy group of
employees who do not need to take much time off work.
However, this low absenteeism trajectory may represent a group
of employees who are less inclined to take time off work when
they are unwell. This is consistent with the concept of presentee-
ism, which can be defined as employees attending work when they
are unwell and should be absent from work (Aronsson & Gustafs-
son, 2005; Dew, Keefe, & Small, 2005; Johns, 2010). Presentee-
ism is associated with productivity losses that are estimated to
outweigh losses associated with absenteeism (Cooper & Dewe,
2008; Hemp, 2004). A range of factors have the potential to
contribute to presenteeism, including organizational policies relat-
ing to sick leave and downsizing, ease of replacement, perceived
pressure for supervisors or coworkers to attend work while unwell,
and low levels of job security (Aronsson & Gustafsson, 2005;
Johns, 2011). The present study did not include a measure of
presenteeism, so it is not possible to determine whether some
employees in this trajectory had higher levels of presenteeism.
However, the findings for job type, long work hours, and self-
employed/business owner status are consistent with existing the-
ories and findings relating to presenteeism. Therefore, it is feasible
that certain characteristics of jobs (e.g., job type, self-employed
status) increase the likelihood of low absenteeism and high pre-
senteeism. These possibilities warrant investigation in future re-
search.
No Absenteeism Trajectory
Employees in the no absenteeism trajectory reported very low
rates of absenteeism annually over the 5-year period, suggesting
 ABSENTEEISM IN FULL-TIME AUSTRALIAN EMPLOYEES
they essentially took no sick leave days each year. This is an
interesting finding because it suggests that although rates of ab-
senteeism in Australia range from 4 – 8 days per year, a substantial
proportion of full-time employees (approximately 25% in this
sample) take very few (if any) sick days each year.
Similar to the low absenteeism trajectory, these individuals were
more likely to be male, and have higher levels of general health
relative to the moderate absenteeism trajectory. As noted above,
these factors may contribute to lower rates of absenteeism. How-
ever, these individuals were younger, which is an interesting
finding given that studies generally find that older (not younger)
age is linked with lower absenteeism (Martocchio, 1989; Ng &
Feldman, 2008). The reasons for this divergent finding are not
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
clear and require further investigation.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
Employees in the no absenteeism trajectory had higher levels of
job control, suggesting more autonomy and flexibility; previous
research suggests that more job control could benefit health and
well-being by allowing individuals to better balance their work and
nonwork commitments, and recover from work stress (Ala-
Mursula et al., 2006). It is therefore feasible that higher levels of
job control could contribute to lower rates of absenteeism over
time (Nielsen et al., 2006; Smulders & Nijhuis, 1999; Väänänen et
al., 2003; Virtanen et al., 2007).
However, other findings suggest that this trajectory could be
characterized, at least to some extent, by presenteeism in a similar
manner to the low absenteeism trajectory. For example, employees
in this trajectory had higher odds of being self-employed or a
business owner. As noted above, they may be less inclined to take
time off work due to financial concerns or loss of business (Bena-
vides et al., 2000). Employees in this trajectory were also more
likely to be mangers. Consistent with the low absenteeism trajec-
tory and Nicholson and Johns’ (1985) theory, some managers may
avoid taking time off work when unwell due to perceptions of sick
leave as not being legitimate.
These findings in combination with those for low absenteeism
trajectory, suggest caution when interpreting levels of absentee-
ism. It is possible that some employees are healthier or have better
control over the working arrangements and thus do not need to
take as much time off work. This could translate into low rates of
absenteeism over time. However, low absenteeism is not neces-
sarily an indicator of healthy and productive employees. This is
because some employees may feel pressure to avoid taking time
off work when they are unwell because of workplace cultural
factors (e.g., perceptions of absence as less legitimate) or financial
reasons (e.g., loss of income for self-employees). This may indi-
cate higher rates of presenteeism which also has the potential to
contribute to lost productivity, perhaps at levels higher than ab-
senteeism (Hemp, 2004).
Strengths and Limitations
The present study provides an important extension on existing
absenteeism literature. First, the sophisticated modeling approach
to examine longitudinal data across five time points provided
insights into intraindividual and interindividual differences in pat-
terns of absenteeism over time. This supports a small number of
studies that have recently identified distinct longitudinal trajecto-
ries of absenteeism (Dello Russo, et al., 2013; Haukka et al., 2014;
Haukka et al., 2013). The sample size was also relatively large,
33
providing sufficient statistical power to identify and distinguish
between the four absenteeism trajectories. In addition, we were
able to examine whether demographic, work and job characteris-
tics, and health status differentiated between these groups. This is
important because well-established theories of absenteeism, such
as Steers and Rhodes’ (1978) model, suggest that these factors are
important predictors of absenteeism behaviors. Thus, our results
provide insight into some of the factors that may predict certain
patterns of absenteeism over time.
There are some limitations of the present study that warrant
discussion. First, absenteeism data were derived from self-reported
retrospective recall of absenteeism over a 12-month period. Self-
report measures are widely used but can lead to underestimates of
absenteeism; this may explain why the rates of absenteeism re-
ported in this paper were lower than studies using records-based
data (Johns & Miraglia, 2015). However, Johns and Miraglia
(2015) indicated that absenteeism data from self-report measures
have reasonable rank order convergence with data obtained
through more objective measures (e.g., organizational records).
Therefore, although the identified trajectories may not provide a
precise indication of absenteeism amount, they are likely to accu-
rately capture patterns of absenteeism over time.
Another consideration is that research has demonstrated that the
frequency and duration of absenteeism spells is important in dis-
tinguishing between voluntary and involuntary forms of absentee-
ism (Bakker, Demerouti, de Boer, & Schaufeli, 2003). For exam-
ple, more frequent spells of absenteeism could reflect motivational
factors (such as low job satisfaction), whereas longer spells are due
to health impairment (Bakker et al., 2003). Future research cap-
turing frequency and duration, as well as the underlying reasons
for absences will be important in further delineating the nature of
longitudinal trajectories.
The present study also focused on absenteeism only. In order to
make clearer conclusions about the implications of each trajectory
for employee productivity, future research will need to explore
other components of productivity such as presenteeism. This is
particularly important for the low absenteeism trajectories, where
lower rates of absenteeism could potentially be offset by higher
rates of presenteeism.
There are some other issues that require consideration. A small
proportion of participants (15.4% of the initial, eligible sample)
were excluded due to missing absenteeism data. The characteris-
tics of these participants differed significantly from those who
were included in the final sample. Most notably, excluded partic-
ipants had poorer self-reported health, lower job security, and were
more likely to be self-employed compared with included partici-
pants. As discussed above, these factors have important implica-
tions for patterns of sick leave over time and thus it is feasible that
excluded individuals could represent a unique subgroup of em-
ployees who have differing patterns of sick leave over time. Thus
a potential limitation of this paper is that the exclusion of individ-
uals due to missing data could have implications for the nature and
size of the identified trajectories.
The present paper also focused on full-time employees, and it is
possible that employees who work part-time hours have very
different patterns of absenteeism reflecting differing antecedents.
Part-time employees are an important population, with existing
research demonstrating differences between part-time and full-
time employees in relation to a number of domains such as job
 34 MAGEE, CAPUTI, AND LEE
involvement; these associations vary depending on specific work
arrangements (e.g., permanent/temporary, voluntary/involuntary;
Thorsteinson, 2003). These and other factors may lead to differing
patterns of absenteeism reflecting differing antecedents compared
with full-time employees. We propose that the unique character-
istics and considerable heterogeneity of part-time employees war-
rant separate investigations to fully understand trajectories of
absenteeism and the associated antecedents. Therefore, it is sug-
gested that in addition to clarifying the nature of distinct absen-
teeism trajectories in full-time employees, future studies are also
conducted to explore absenteeism trajectories and their anteced-
ents in part-time employees.
Furthermore, some of the potential antecedents assessed in this
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
study were assessed using short-scales or single items scales (e.g.,
This document is copyrighted by the American Psychological Association or one of its allied publishers.
job satisfaction). This is a potential limitation because these mea-
sures may not provide a comprehensive insight into these factors.
Despite this, single-item scales generally provide appropriate and
valid measures of these constructs and correspond well with mul-
tiple item scales (Wanous, Reichers, & Hudy, 1997). Factors not
examined in this paper such as insomnia, work-life balance, and
organization factors such as specific policies and cultural factors
regarding sick leave and leadership could also be important pre-
dictors of these trajectories and require investigation in future
research.
Conclusions
This paper provides a novel insight into different patterns of
absenteeism over time in a large, heterogonous sample of Austra-
lian employees. The results demonstrate that several sociodemo-
graphic, health, and work-related factors underlie these trajecto-
ries, which is consistent with existing theories of absenteeism. The
pattern of results for the high absenteeism trajectory suggests that
poor health may constrain an employee’s ability to attend work.
The findings for low absenteeism and no absenteeism suggest a
range of potential underlying factors, which may reflect a greater
ability and motivation to attend work or higher levels of presen-
teeism. These findings require further investigation, with measures
of presenteeism needed to better understand these trajectories.
Nevertheless, these results have important implications for em-
ployers and organizations, as they could inform strategies to min-
imize the effects of absenteeism on lost workplace productivity.
These may include providing employees with more flexible work
conditions (e.g., options to work from home, or have variable start
and finish times), greater autonomy, and improved coping skills.
Although there is a need to limit avoidable absenteeism (e.g.,
through prevention strategies), it is also important that organiza-
tions foster workplace cultures that encourage employees to take
sick leave when they are unwell to prevent presenteeism which
could ultimately lead to greater productivity losses.
References
Ala-Mursula, L., Vahtera, J., Kouvonen, A., Väänänen, A., Linna, A.,
Pentti, J., & Kivimäki, M. (2006). Long hours in paid and domestic work
and subsequent sickness absence: Does control over daily working hours
matter? Occupational and Environmental Medicine, 63, 608 – 616.
http://dx.doi.org/10.1136/oem.2005.023937
Aronsson, G., & Gustafsson, K. (2005). Sickness presenteeism: Preva-
lence, attendance-pressure factors, and an outline of a model for re-
search. Journal of Occupational and Environmental Medicine, 47, 958 –
966. http://dx.doi.org/10.1097/01.jom.0000177219.75677.17
Audit Office of New South Wales. (2010). Sick leave: Department of
Premier and Cabinet. Sydney, Australia: Audit Office of New South
Wales.
Australian Bureau of Statistics. (2006). Australian and New Zealand stan-
dard classification of occupations (1st ed.). Canberra, Australia: Aus-
tralian Bureau of Statistics.
Bakker, A. B., Demerouti, E., de Boer, E., & Schaufeli, W. B. (2003). Job
demands and job resources as predictors of absence duration and fre-
quency. Journal of Vocational Behavior, 62, 341–356. http://dx.doi.org/
10.1016/S0001-8791(02)00030-1
Beemsterboer, W., Stewart, R., Groothoff, J., & Nijhuis, F. (2009). A
literature review on sick leave determinants (1984 –2004). International
Journal of Occupational Medicine and Environmental Health, 22, 169 –
179. http://dx.doi.org/10.2478/v10001-009-0013-8
Benavides, F. G., Benach, J., Diez-Roux, A. V., & Roman, C. (2000). How
do types of employment relate to health indicators? Findings from the
second European survey on working conditions. Journal of Epidemiol-
ogy and Community Health, 54, 494 –501. http://dx.doi.org/10.1136/jech
.54.7.494
Blekesaune, M. (2012). Job insecurity and sickness absence: Correlations
between attrition and absence in 36 occupational groups. Scandinavian
Journal of Public Health, 40, 668 – 673. http://dx.doi.org/10.1177/
1403494812459816
Brooke, P. P., Jr. (1986). Beyond the Steers and Rhodes model of em-
ployee attendance. The Academy of Management Review, 11, 345–361.
Confederation of British Industry. (2004). Absence and labour turnover
survey 2003. The lost billions: Addressing the cost of absence. London,
UK: Confederation of British Industry.
Cooper, C., & Dewe, P. (2008). Well-being: Absenteeism, presenteeism,
costs and challenges. Occupational Medicine, 58, 522–524. http://dx.doi
.org/10.1093/occmed/kqn124
Darr, W., & Johns, G. (2008). Work strain, health, and absenteeism: A
meta-analysis. Journal of Occupational Health Psychology, 13, 293–
318. http://dx.doi.org/10.1037/a0012639
Dello Russo, S., Miraglia, M., Borgogni, L., & Johns, G. (2013). How time
and perceptions of social context shape employee absenteeism trajecto-
ries. Journal of Vocational Behavior, 83, 209 –217. http://dx.doi.org/
10.1016/j.jvb.2013.03.005
Dew, K., Keefe, V., & Small, K. (2005). “Choosing” to work when sick:
Workplace presenteeism. Social Science & Medicine, 60, 2273–2282.
http://dx.doi.org/10.1016/j.socscimed.2004.10.022
Fair Work Ombudsman. (2010). Sick and carer’s leave. Retrieved from
http://www.fairwork.gov.au/leave/sick-and-carers-leave
Feeney, A., North, F., Head, J., Canner, R., & Marmot, M. (1998).
Socioeconomic and sex differentials in reason for sickness absence from
the Whitehall II Study. Occupational and Environmental Medicine, 55,
91–98. http://dx.doi.org/10.1136/oem.55.2.91
Fekedulegn, D., Burchfiel, C. M., Hartley, T. A., Andrew, M. E., Charles,
L. E., Tinney-Zara, C. A., & Violanti, J. M. (2013). Shiftwork and
sickness absence among police officers: The BCOPS study. Chronobi-
ology International, 30, 930 –941. http://dx.doi.org/10.3109/07420528
.2013.790043
Goetzel, R. Z., Long, S. R., Ozminkowski, R. J., Hawkins, K., Wang, S.,
& Lynch, W. (2004). Health, absence, disability, and presenteeism cost
estimates of certain physical and mental health conditions affecting U.S.
employers. Journal of Occupational and Environmental Medicine, 46,
398 – 412. http://dx.doi.org/10.1097/01.jom.0000121151.40413.bd
Halbesleben, J. R. B., Whitman, M. V., & Crawford, W. S. (2014). A
dialectical theory of the decision to go to work: Bringing together
absenteeism and presenteeism. Human Resource Management Review,
24, 177–192. http://dx.doi.org/10.1016/j.hrmr.2013.09.001
 ABSENTEEISM IN FULL-TIME AUSTRALIAN EMPLOYEES
Hansson, A. S., Vingård, E., Arnetz, B. B., & Anderzén, I. (2008).
Organizational change, health, and sick leave among health care em-
ployees: A longitudinal study measuring stress markers, individual, and
work site factors. Work & Stress, 22, 69 – 80. http://dx.doi.org/10.1080/
02678370801996236
Hardy, G. E., Woods, D., & Wall, T. D. (2003). The impact of psycho-
logical distress on absence from work. Journal of Applied Psychology,
88, 306 –314. http://dx.doi.org/10.1037/0021-9010.88.2.306
Haukka, E., Kaila-Kangas, L., Luukkonen, R., Takala, E. P., Viikari-
Juntura, E., & Leino-Arjas, P. (2014). Predictors of sickness absence
related to musculoskeletal pain: A two-year follow-up study of workers
in municipal kitchens. Scandinavian Journal of Work, Environment &
Health, 40, 278 –286. http://dx.doi.org/10.5271/sjweh.3415
Haukka, E., Kaila-Kangas, L., Ojajärvi, A., Miranda, H., Karppinen, J.,
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Viikari-Juntura, E., . . . Leino-Arjas, P. (2013). Pain in multiple sites and
This document is copyrighted by the American Psychological Association or one of its allied publishers.
sickness absence trajectories: A prospective study among Finns. Pain,
154, 306 –312. http://dx.doi.org/10.1016/j.pain.2012.11.003
Head, J., Kivimäki, M., Martikainen, P., Vahtera, J., Ferrie, J. E., &
Marmot, M. G. (2006). Influence of change in psychosocial work
characteristics on sickness absence: The Whitehall II Study. Journal of
Epidemiology and Community Health, 60, 55– 61. http://dx.doi.org/
10.1136/jech.2005.038752
Hemp, P. (2004). Presenteeism: At work—But out of it. Harvard Business
Review, 82, 49 –58, 155.
Johns, G. (2008). Absenteeism and presenteeism: Not at work or not
working well. In C. L. Cooper & J. Barling (Eds.), The Sage handbook
of organizational behavior (Vol. 1, pp. 160 –177). London: Sage.
Johns, G. (2010). Presenteeism in the workplace: A review and research
agenda. Journal of Organizational Behavior, 31, 519 –542. http://dx.doi
.org/10.1002/job.630
Johns, G. (2011). Attendance dynamics at work: The antecedents and
correlates of presenteeism, absenteeism, and productivity loss. Journal
of Occupational Health Psychology, 16, 483–500. http://dx.doi.org/
10.1037/a0025153
Johns, G., & Miraglia, M. (2015). The reliability, validity, and accuracy of
self-reported absenteeism from work: A meta-analysis. Journal of Oc-
cupational Health Psychology, 20, 1–14.
Jung, T., & Wickrama, K. A. S. (2008). An introduction to latent class
growth analysis and growth mixture modeling. Social and Personality
Psychology Compass, 2, 302–317. http://dx.doi.org/10.1111/j.1751-
9004.2007.00054.x
Kääriä, S., Laaksonen, M., Leino-Arjas, P., Saastamoinen, P., & Lahelma,
E. (2012). Low back pain and neck pain as predictors of sickness
absence among municipal employees. Scandinavian Journal of Public
Health, 40, 150 –156. http://dx.doi.org/10.1177/1403494811435490
Karasek, R. A. (1979). Job demands, job decision latitude and mental
strain: Implications for job redesign. Administrative Science Quarterly,
24, 285–308. http://dx.doi.org/10.2307/2392498
Kivimäki, M., Vahtera, J., Thomson, L., Griffiths, A., Cox, T., & Pentti, J.
(1997). Psychosocial factors predicting employee sickness absence dur-
ing economic decline. Journal of Applied Psychology, 82, 858 – 872.
http://dx.doi.org/10.1037/0021-9010.82.6.858
Knudsen, A. K., Harvey, S. B., Mykletun, A., & Øverland, S. (2013).
Common mental disorders and long-term sickness absence in a general
working population. The Hordaland Health Study. Acta Psychiatrica
Scandinavica, 127, 287–297. http://dx.doi.org/10.1111/j.1600-0447
.2012.01902.x
Labriola, M., Lund, T., & Burr, H. (2006). Prospective study of physical
and psychosocial risk factors for sickness absence. Occupational Med-
icine, 56, 469 – 474. http://dx.doi.org/10.1093/occmed/kql058
Magee, C., Stefanic, N., Caputi, P., & Iverson, D. (2011). Occupational
factors and sick leave in Australian employees. Journal of Occupational
and Environmental Medicine, 53, 627– 632. http://dx.doi.org/10.1097/
JOM.0b013e31821df4cf
35
Martocchio, J. J. (1989). Age-related differences in employee absenteeism:
A meta-analysis. Psychology and Aging, 4, 409 – 414. http://dx.doi.org/
10.1037/0882-7974.4.4.409
Mastekaasa, A., & Olsen, K. M. (1998). Gender, absenteeism, and job
characteristics: A fixed effects approach. Work and Occupations, 25,
195–228. http://dx.doi.org/10.1177/0730888498025002004
McTernan, W. P., Dollard, M. F., & LaMontagne, A. D. (2013). Depres-
sion in the workplace: An economic cost analysis of depression-related
productivity loss attributable to job strain and bullying. Work & Stress,
27, 321–338. http://dx.doi.org/10.1080/02678373.2013.846948
Medibank Private. (2005). The health of Australia’s workforce. Melbourne,
Australia: Medibank Private.
Melchior, M., Ferrie, J. E., Alexanderson, K., Goldberg, M., Kivimäki, M.,
Singh-Manoux, A., . . . Head, J. (2009). Using sickness absence records
to predict future depression in a working population: Prospective find-
ings from the GAZEL cohort. American Journal of Public Health, 99,
1417–1422. http://dx.doi.org/10.2105/AJPH.2008.142273
Melchior, M., Niedhammer, I., Berkman, L. F., & Goldberg, M. (2003). Do
psychosocial work factors and social relations exert independent effects
on sickness absence? A six year prospective study of the GAZEL cohort.
Journal of Epidemiology and Community Health, 57, 285–293. http://
dx.doi.org/10.1136/jech.57.4.285
Muthén, B. O. (2004). Latent variable analysis: Growth mixture modeling
and related techniques for longitudinal data. In D. Kaplan (Ed.), Hand-
book of quantitative methodology for the social sciences (pp. 346 –368).
Newbury Park, CA: Sage. http://dx.doi.org/10.4135/9781412986311
.n19
Muthén, B. (2006). The potential of growth mixture modelling. Infant and
Child Development, 15, 623– 625. http://dx.doi.org/10.1002/icd.482
Muthén, L. K., & Muthén, B. O. (1998 –2010). Mplus users guide. Sixth
Edition. Los Angeles: CA: Muthén & Muthén.
Ng, T. W. H., & Feldman, D. C. (2008). The relationship of age to ten
dimensions of job performance. Journal of Applied Psychology, 93,
392– 423. http://dx.doi.org/10.1037/0021-9010.93.2.392
Nicholson, N. (1977). Absence behaviour and attendance motivation: A
conceptual synthesis. Journal of Management Studies, 14, 231–252.
http://dx.doi.org/10.1111/j.1467-6486.1977.tb00364.x
Nicholson, N., & Johns, G. (1985). The absence culture and the psycho-
logical contract: Who’s in control of absence? The Academy of Man-
agement Review, 10, 397– 407.
Nielsen, M. L., Rugulies, R., Christensen, K. B., Smith-Hansen, L., &
Kristensen, T. S. (2006). Psychosocial work environment predictors of
short and long spells of registered sickness absence during a 2-year
follow up. Journal of Occupational and Environmental Medicine, 48,
591–598. http://dx.doi.org/10.1097/01.jom.0000201567.70084.3a
Nylund, K. L., Asparouhov, T., & Muthén, B. O. (2007). Deciding on the
number of classes in latent class analysis and growth mixture modeling:
A Monte Carlo simulation study. Structural Equation Modeling, 14,
535–569. http://dx.doi.org/10.1080/10705510701575396
Pousette, A., & Hanse, J. J. (2002). Job characteristics as predictors of
ill-health and sickness absenteeism in different occupational types–a
multigroup structural equation modelling approach. Work & Stress, 16,
229 –250. http://dx.doi.org/10.1080/02678370210162737
Roelen, C. A. M., Hoedeman, R., van Rhenen, W., Groothoff, J. W., van
der Klink, J. J., & Bültmann, U. (2014). Mental health symptoms as
prognostic risk markers of all-cause and psychiatric sickness absence in
office workers. European Journal of Public Health, 24, 101–105. http://
dx.doi.org/10.1093/eurpub/ckt034
Rugulies, R., Christensen, K. B., Borritz, M., Villadsen, E., Bültmann, U.,
& Kristensen, T. S. (2007). The contribution of the psychosocial work
environment to sickness absence in human service workers: Results of a
3-year follow-up study. Work & Stress, 21, 293–311. http://dx.doi.org/
10.1080/02678370701747549
 36 MAGEE, CAPUTI, AND LEE

Schalk, R. (2011). The influence of organizational commitment and health
on sickness absenteeism: A longitudinal study. Journal of Nursing
Management, 19, 596 – 600. http://dx.doi.org/10.1111/j.1365-2834.2010
.01170.x
Schaufeli, W. B., Bakker, A. B., & Van Rhenen, W. (2009). How changes
in job demands and resources predict burnout, work engagement, and
sickness absenteeism. Journal of Organizational Behavior, 30, 893–917.
http://dx.doi.org/10.1002/job.595
Siegrist, J. (1996). Adverse health effects of high-effort/low-reward con-
ditions. Journal of Occupational Health Psychology, 1, 27– 41. http://
dx.doi.org/10.1037/1076-8998.1.1.27
Smulders, P. G., & Nijhuis, F. J. N. (1999). The job demands-job
control model and absence behaviour: Results of a 3-year longitudi-
nal study. Work & Stress, 13, 115–131. http://dx.doi.org/10.1080/
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Vahtera, J., Kivimäki, M., Pentti, J., & Theorell, T. (2000). Effect of
change in the psychosocial work environment on sickness absence: A
seven year follow up of initially healthy employees. Journal of Epide-
miology and Community Health, 54, 484 – 493. http://dx.doi.org/
10.1136/jech.54.7.484
Virtanen, M., Vahtera, J., Pentti, J., Honkonen, T., Elovainio, M., &
Kivimäki, M. (2007). Job strain and psychologic distress influence on
sickness absence among Finnish employees. American Journal of Pre-
ventive Medicine, 33, 182–187. http://dx.doi.org/10.1016/j.amepre.2007
.05.003
Wada, K., Arakida, M., Watanabe, R., Negishi, M., Sato, J., & Tsutsumi,
A. (2013). The economic impact of loss of performance due to absen-
teeism and presenteeism caused by depressive symptoms and comorbid
health conditions among Japanese workers. Industrial Health, 51, 482–

026783799296093 489. http://dx.doi.org/10.2486/indhealth.2013-0016

Wanous, J. P., Reichers, A. E., & Hudy, M. J. (1997). Overall job
This document is copyrighted by the American Psychological Association or one of its allied publishers.

Steers, R. M., & Rhodes, S. R. (1978). Major influences on employee

attendance: A process model. Journal of Applied Psychology, 63, 391–
407. http://dx.doi.org/10.1037/0021-9010.63.4.391
Suominen, S., Vahtera, J., Korkeila, K., Helenius, H., Kivimäki, M., &
Koskenvuo, M. (2007). Job strain, life events, and sickness absence: A
longitudinal cohort study in a random population sample. Journal of
Occupational and Environmental Medicine, 49, 990 –996. http://dx.doi
.org/10.1097/JOM.0b013e3181343e2b
ten Brummelhuis, L. L., Ter Hoeven, C. L., De Jong, M. D. T., & Peper,
B. (2013). Exploring the linkage between the home domain and absence
from work: Health, motivation, or both? Journal of Organizational
Behavior, 34, 273–290. http://dx.doi.org/10.1002/job.1789
Thorsteinson, T. J. (2003). Job attitudes of part-time vs. full-time
workers: A meta-analytic review. Journal of Occupational and Or-
ganizational Psychology, 76, 151–177. http://dx.doi.org/10.1348/
096317903765913687
Väänänen, A., Toppinen-Tanner, S., Kalimo, R., Mutanen, P., Vahtera, J.,
& Peiró, J. M. (2003). Job characteristics, physical and psychological
symptoms, and social support as antecedents of sickness absence among
men and women in the private industrial sector. Social Science &
Medicine, 57, 807– 824. http://dx.doi.org/10.1016/S0277-9536
(02)00450-1
satisfaction: How good are single-item measures? Journal of Applied
Psychology, 82, 247–252. http://dx.doi.org/10.1037/0021-9010.82.2.247
Ware, J. E., Kosinski, M., & Gandek, B. (2000). SF-36 Health Survey
manual and interpretation guide. Lincoln, RI: Quality Metric Incorpo-
rated.
Westman, M., & Etzion, D. (2001). The impact of vacation and job stress
on burnout and absenteeism. Psychology & Health, 16, 595– 606. http://
dx.doi.org/10.1080/08870440108405529
Wooden, M., Freidin, S., & Watson, N. (2002). Household, Income and
Labour Dynamics in Australia (HILDA) Survey: Wave 1. The Austra-
lian Economic Review, 35, 339 –348. http://dx.doi.org/10.1111/1467-
8462.00252
Ybema, J. F., Smulders, P. G. W., & Bongers, P. M. (2010). Antecedents
and consequences of employee absenteeism: A longitudinal perspective
on the role of job satisfaction and burnout. European Journal of Work
and Organizational Psychology, 19, 102–124. http://dx.doi.org/10.1080/
13594320902793691
Received April 27, 2014
Revision received February 3, 2015
Accepted February 16, 2015 䡲

E-Mail Notification of Your Latest Issue Online!

Would you like to know when the next issue of your favorite APA journal will be available
online? This service is now available to you. Sign up at http://notify.apa.org/ and you will be
notified by e-mail when issues of interest to you become available!